- Rating
- 1
- Confidence
- 3
- Informativeness
- 1
- Presentation
- 3
- Soundness
- 1
Read review
Paper summary
WaypointVLA focuses on the efficiency of VLA models. It predicts sparse waypoints and then uses interpolation to produce an action sequence, achieving a large speedup. However, this approach may reduce generalization and robustness, and it does not appear to scale well.
Critical strengths
-
Efficiency is a key issue for VLA models. The paper explores a design for the action expert that generates trajectories through sparse prediction followed by interpolation.
-
The paper includes a wide range of robustness tests under different environmental conditions.
-
The writing and figures are easy to follow.
Critical weaknesses
-
The sparse prediction approach does not show that longer-horizon prediction is unnecessary. Predicting over longer horizons is essential for more complex tasks, but the paper does not evaluate its method on such tasks.
-
Interpolation can work well for simple tasks in relatively static environments. However, the method may struggle when the environment changes or when unexpected events during execution require corrections. The robustness evaluation should focus on changes during task execution, rather than only testing different static environment settings.
-
The Duration Head is trained on the full dataset of 1.4k samples, while the paper claims to use only 200 demonstrations. To support this claim, all components should be trained using only those 200 demonstrations, and the comparisons should use the same data budget.
Questions for authors
Please refer to the weaknesses section. I am willing to raise the score if the authors can address my concerns.
4 comments
W2. Robustness to dynamic perturbations during execution.
A2: We thank the reviewer for this valuable suggestion. In our original experiments, we evaluated WaypointVLA and the dense policy on LIBERO-Plus under seven types of perturbations. As shown below, WaypointVLA consistently outperforms the dense baseline across all demonstration budgets, demonstrating improved robustness to visual and environmental changes.
| Method | Episodes | Camera | Robot | Language | Light | Background | Noise | Layout | Total |
|---|---|---|---|---|---|---|---|---|---|
| Dense | 200 | 53.1 | 63.5 | 71.3 | 80.0 | 80.4 | 68.0 | 73.3 | 68.9 |
| WaypointVLA | 200 | 65.2 | 70.3 | 76.8 | 89.3 | 83.3 | 79.2 | 78.6 | 76.7 |
| Dense | 400 | 55.7 | 66.7 | 71.7 | 81.0 | 78.9 | 71.4 | 73.8 | 70.5 |
| WaypointVLA | 400 | 68.8 | 72.1 | 77.9 | 85.1 | 85.7 | 77.9 | 80.7 | 77.6 |
| Dense | 600 | 58.2 | 66.6 | 73.9 | 84.3 | 81.4 | 74.2 | 73.9 | 72.3 |
| WaypointVLA | 600 | 69.0 | 74.9 | 79.4 | 89.4 | 89.1 | 80.0 | 82.7 | 79.8 |
| Dense | 1,439 | 56.3 | 67.6 | 77.3 | 83.4 | 83.6 | 74.8 | 75.5 | 73.2 |
| WaypointVLA | 1,439 | 73.3 | 75.3 | 82.8 | 95.3 | 93.7 | 85.4 | 84.0 | 83.3 |
To address the reviewer's concern that these perturbations are primarily static, we further conducted experiments with perturbations introduced during task execution, in both real-world and simulated environments.
1. Real-world dynamic lighting perturbations. In our real-world experiments, we randomly changed the environmental illumination using a handheld flashlight while the robot was executing the task. Each model was independently evaluated over 100 trials. Under these dynamic lighting changes, WaypointVLA achieved 82% success, compared with 74% for the dense policy, showing an 8-percentage-point improvement.
2. Dynamic perturbations in LIBERO. We additionally evaluated both models on the 10 LIBERO-Goal tasks with Light, Background, and Noise perturbations introduced during execution. Specifically, perturbations were randomly applied at selected time points (0.3, 1.0, 1.7, and 2.4 seconds), each lasting 0.5 seconds. We varied the number, combinations, and temporal order of perturbations to simulate changing visual conditions during task execution. To ensure a fair comparison, both models were evaluated under matched perturbation distributions, with 100 trials per perturbation category for each model.
| Method | Light | Background | Noise | Overall |
|---|---|---|---|---|
| Dense FT | 91.0 | 94.0 | 92.0 | 92.33 |
| WaypointVLA | 95.0 | 94.0 | 94.0 | 94.33 |
WaypointVLA achieves 94.33% overall success, compared with 92.33% for the dense baseline, improving performance by 2.0 percentage points. It achieves higher success rates under dynamic lighting and noise perturbations, while maintaining comparable performance under background changes.
These additional results suggest that WaypointVLA remains effective when visual perturbations occur during task execution, rather than only under static environmental changes. While the current experiments focus on visual disturbances, they provide further evidence that sparse waypoint prediction and interpolation can maintain robust performance under dynamically changing observations.
W3. Data-budget mismatch in Duration Head training
A3: We thank the reviewer for identifying this issue. We acknowledge that sharing a Duration Head trained on the full 1,439 demonstrations across different low-data settings was an oversight in our experimental protocol.
To address this concern, we retrained separate Duration Heads for the 200-, 400-, and 600-demonstration settings, ensuring that each Duration Head only uses samples derived from the corresponding training subset. We have also updated the corresponding results, tables, and figures in the revised manuscript accordingly.
We then reevaluated the updated models on both LIBERO and LIBERO-Plus.
1. Updated LIBERO-Plus results.
| Episodes | Method | Camera | Robot | Language | Light | Background | Noise | Layout | Total |
|---|---|---|---|---|---|---|---|---|---|
| 200 | Dense | 53.1 | 63.5 | 71.3 | 80.0 | 80.4 | 68.0 | 73.3 | 68.9 |
| Waypt. | 65.2 | 70.3 | 76.8 | 89.3 | 83.3 | 79.2 | 78.6 | 76.7 | |
| 400 | Dense | 55.7 | 66.7 | 71.7 | 81.0 | 78.9 | 71.4 | 73.8 | 70.5 |
| Waypt. | 68.8 | 72.1 | 77.9 | 85.1 | 85.7 | 77.9 | 80.7 | 77.6 | |
| 600 | Dense | 58.2 | 66.6 | 73.9 | 84.3 | 81.4 | 74.2 | 73.9 | 72.3 |
| Waypt. | 69.0 | 74.9 | 79.4 | 89.4 | 89.1 | 80.0 | 82.7 | 79.8 | |
| 1,439 | Dense | 56.3 | 67.6 | 77.3 | 83.4 | 83.6 | 74.8 | 75.5 | 73.2 |
| Waypt. | 73.3 | 75.3 | 82.8 | 95.3 | 93.7 | 85.4 | 84.0 | 83.3 |
Compared with the original results, the budget-matched Duration Heads lead to some performance decreases, particularly in the low-data setting:
| Episodes | Original Total (%) | Updated Total (%) | Δ (pp) |
|---|---|---|---|
| 200 | 78.90 | 76.72 | −2.18 |
| 400 | 78.90 | 77.63 | −1.27 |
| 600 | 80.00 | 79.83 | −0.17 |
Nevertheless, WaypointVLA still consistently outperforms the corresponding dense baseline across all demonstration budgets, achieving improvements of +7.8, +7.1, and +7.5 percentage points with 200, 400, and 600 demonstrations, respectively.
2. Updated LIBERO results.
We also reevaluated the models on LIBERO using the budget-matched Duration Heads:
| Episodes | Original Best SR (%) | Updated Best SR (%) | Δ (pp) |
|---|---|---|---|
| 200 | 91.50 | 90.25 | −1.25 |
| 400 | 94.00 | 93.50 | −0.50 |
| 600 | 94.00 | 94.50 | +0.50 |
We observe that the performance difference generally decreases as the training budget increases, suggesting that Duration Head estimation benefits from additional demonstrations.
More importantly, even when the Duration Head is trained using only 200 demonstrations, WaypointVLA achieves 90.25% on LIBERO and 76.72% on LIBERO-Plus, outperforming the corresponding dense baselines (88.25% and 68.9%, respectively).
These additional experiments confirm that the data-efficiency advantage of WaypointVLA remains under strictly matched demonstration budgets, without requiring the Duration Head to access additional demonstrations.
Polite follow-up: request for reviewers’ feedback on our responses!
Hi, we sincerely thank you very much for these constructive comments and evaluation of our manuscript. We would like to kindly ask you to take a look at our responses and reevaluate our work based on our clarifications. Please let us know whether our response addresses your concerns or whether there is any further detail we can provide to help address these concerns.
Thank you again for dedicating your time to reviewing our paper.
W1. Long-horizon prediction.
A1: We thank the reviewer for this valuable suggestion. We would like to clarify that our method does not claim that longer-horizon prediction is unnecessary, and we have clarified this point in the revised abstract. Instead, our primary motivation is to reduce redundant training costs caused by the prediction–execution mismatch in dense action prediction. Our key observation is that intermediate actions within relatively smooth segments can be reconstructed through linear interpolation. Moreover, WaypointVLA employs an adaptive Duration Head, allowing the execution horizon to vary rather than being restricted to a fixed short duration.
1. Adaptive Duration Analysis. To examine this adaptive behavior, we analyzed the predicted duration distribution during LIBERO evaluation. The results are summarized below.
We observe that LIBERO-Long exhibits a larger average predicted duration (3.91 steps) than the other three suites (2.73–2.91 steps), with a 95th percentile of 6 steps. This indicates that WaypointVLA can adaptively predict longer execution segments for long-horizon tasks rather than always relying on short, fixed-duration predictions.
2. Long-horizon Task Evaluation. To further address the reviewer's concern, we conducted a task-level comparison between WaypointVLA and the original π0.5 dense policy on all LIBERO-Long tasks.
WaypointVLA achieves 89.0% overall success, compared with 79.0% for the original π0.5 dense policy, showing a 10.0-percentage-point improvement. In particular, WaypointVLA achieves higher success rates on 6 of the 10 tasks, comparable performance on 2 tasks, and lower performance on the remaining 2 tasks.
These results suggest that sparse waypoint prediction can remain effective on multi-stage, long-horizon manipulation tasks. Rather than eliminating long-horizon reasoning or control, our method reduces redundant dense action supervision while allowing the execution duration to adapt to different segments.